Complexos miméticos em cobras-corais em uma região do Pantanal Norte (Fazenda Retiro Novo, Poconé - MT) e em uma área de Cerrado (Parque Nacional Chapada dos Guimarães - MT)

Abstract

Since the XIXth century, the existence of mimicry complexes among poisonous and non- poisonous coralsnakes in the Neotropical region has stimulated controversial discussions

in the literature. In this context, experimental studies employing plasticine replicas of the several coralsnake patterns represent classical approaches aiming to quantify and compare predation rates by different kinds of predators. In addition, the method allows identifying predator categories (e.g., birds and mammals) based on the general aspects of the jaw prints left on the replicas at each predation attempt. Although most of these works have corroborated the protective character of coralsnakes color, the literature lacks studies focused on the problem, but taking climatic parameters into account. Herein, I used plasticine replicas to approach coralsnakes mimicry complexes in two different sites, one in the Cerrado Domain, at Chapada dos Guimarães National Park, and the other in the Pantanal floodplain, municipality of Poconé, both in Mato Grosso state, Midwest Brazil. The vegetation of these areas is mostly represented by open formations and the climate is typically seasonal, with the Pantanal exhibiting a flooding cycle that restricts the habitats of terrestrial species for a long period throughout the year. The sampling effort consisted of 760 plasticine replicas representing the color pattern of the most common coralsnake species in each one of the areas, as well as monochromatic (uniform brown) replicas as the control group. The replicas were distributed in the field throughout six field campaigns undertaken in the period of one year. Statistical analyses of the data included Chi-squared tests to perform pairwise comparisons of predation rates among color patterns, and Binomial Logistic Regressions to estimate the effect of coloration in the general predation rates, as well as the probabilities of each one of the color patterns to suffer attacks. My results suggest that (1) birds represent the most frequent predators in both areas, respectively followed by mammals and lizards; (2) in both study sites, in contrast with most previous studies, the monochromatic replicas suffered a sensibly smaller number of attacks by virtually all predator categories than the aposematic replicas; (3) the very low incidence of attacks in the Cerrado of Chapada dos Guimarãesa (3.3%) follows the general patterns of similar studies in the same domain (nearly 6%), and (4) in the Pantanal, the pattern of predation rates is the opposite that those of most previous studies, with the replicas representing the toxic model receiving the largest rates of predation in comparison with all other patterns (non-toxic mimics and monochromatic replicas). Although we consider our results from the Cerrado inconclusive requiring larger sampling efforts, or results in the Pantanal suggest that the flooding regime may lead, not only to higher predation rates, but also to lower selectivity by visually oriented predators. Factors as the high concentrations of terrestrial predators and preys during the flooding season, the relative abundance of toxic models and mimics, and the higher detectability of coralsnakes by predators with accurate vision may help explaining the apparent inefficiency of the establishment of mimicry complexes, at least at local perspective. Additional experiments in other localities of the Pantanal floodplain would be welcome to clarify the problem.